Computer clusters typically include multiple nodes that collectively perform one or more computing tasks. Occasionally, a network-partition event (such as a device or communication-infrastructure failure) may impair communication within a computer cluster, effectively partitioning the computer cluster into multiple sub-clusters (e.g., groups of one or more nodes) that are unable to communicate with each other. Due to the network-partition event, these isolated sub-clusters may be unable to coordinate operations, such as writing data to and/or reading data from shared resources. In order to avoid overwriting and/or corrupting data related to these tasks, system administrators may seek to ensure that only one sub-cluster performs such tasks after a network-partition event occurs.
Multi-cluster computing systems may include various computer clusters that operate independently of each other. When a network-partition event affects computer clusters on a multi-cluster system, the coordination point may allow a single sub-cluster within each computer cluster to perform computing tasks. However, such computer clusters often incorporate sub-clusters that are located at different sites. As a result, traditional multi-cluster systems may sacrifice efficiency and performance by allowing sub-clusters at different sites to perform computing tasks on behalf of the computer clusters. Accordingly, the instant disclosure identifies a need for efficiently and effectively managing sub-clusters within multiple computer clusters subsequent to a network-partition event.